Philip K. Dick’s sci-fi classic “Do Androids Dream of Electric Sheep?” poses a pertinent question. How close can machines come to the fantasies, imperfections and idiosyncrasies of human thought? Oddly enough, imperfection may be our greatest asset. Every facet of the human experience, for better or worse, can inspire written content. For instance, putting too much sugar and cream in your coffee could lead to a metaphor about the need for simplicity in cybersecurity. Or your lengthy commute could produce analogies comparing a highway’s lanes and state troopers to a network’s optical technologies and security solutions.
Experimentation that seems illogical to an algorithm may result in a compelling angle. I’m not trying to sound all “old man yells at cloud in fear of getting booted out of the tech industry” here, but current AI may be too risk-averse to produce simultaneously accurate, original and engaging content. That second adjective brings us to every writer’s nightmare – plagiarism. A Substack recently got caught directly plagiarizing a writer’s work with automated writing tools. On the bright side, the actual writer got to lambast the viral, stolen article with his own viral, original article. Humans warring with AI - it’s “I, Robot” all over again.
The robots are coming...
ChatGPT has made headlines for more positive reasons, showcasing the powerful capabilities of AI for work process automation. The language model interacts through a dialogue format, allowing it to correct itself upon clarification. To be fair, a morbid, unrealistic fixation on perfection is the hallmark of any writer. But what does this mean for technical B2B content? Will I eventually have to study machine learning to serve my ChatGPT overlords? ChatGPT may prove useful in the B2B field for brainstorming blogs or as an alternative search engine, and AI prompt writing will inevitably become an important skill. But can it automate technical content that requires critical thinking, concept differentiation and client-specific messaging? I ran some basic writing prompts to test AI’s writing capability and answer that question. Unfortunately, I resorted to an alternative since ChatGPT’s servers were overloaded - not ideal when you need technical content on short notice! I can almost hear the desperate keyboard clacks of a thousand college students somewhere in the distance.
Putting my computational competitor to the test
Prompt: VPN vs ZTNA and SDP including microsegmentation and device cloaking
There are a few different approaches to securing data, including VPNs, ZTNA, and SDP. Each of these solutions offer their own unique advantages and disadvantages, but one thing they all have in common is the need for microsegmentation. VPNs use microsegmentation to create isolated segments of the network, while ZTNA cloaks devices by masking their IP addresses. All three solutions offer unique advantages when it comes to protecting corporate networks from external threats.
Slow your roll, soldier! VPNs do need microsegmentation, but they generally don’t offer it. Typically, ZTNA and SDP already include microsegmentation. Additionally, ZTNA is the goal of an SDP model, at least for the client I’m writing about. Posing it as a separate solution may confuse customers and shows a lack of contextual understanding of the audience. Maybe it’s the “and SDP” in the prompt.
Let’s dance, Mr. Roboto
Just for fun, let’s add phishing-resistant MFA and peer-to-peer architecture then max out the creativity level. Note that the tool warns of decreased accuracy with increased creativity. Alas, the cost of artistic expression.
Prompt: VPN vs ZTNA with SDP including microsegmentation, phishing-resistant MFA, peer-to-peer architecture and device cloaking
ZTNA with SDP makes use of microsegmentation and device cloaking technologies in order to reduce risks associated with phishing as well as enabling users or applications with phishing-resistant MFA. VPNs and ZTNA with SDP offer an enterprise-grade solutions to insecure remote access. By using techniques such as microsegmentation, device cloaking, phishing-resistant MFA and a peer-to-peer architecture, these solutions provide a secure solution for companies wanting to ensure that their data remains safe. With VPNs and ZTNA with SDP, companies can control who has access to the data they are sharing over the network, as well as where it is stored.
Bit of an awkward first sentence there, computer. Furthermore, I wouldn’t necessarily say phishing-resistant MFA “enables” users or applications. It authenticates users through passwordless methods to provide access to applications. Technical content requires precision. The bot did stop posing ZTNA and SDP as disparate. Unfortunately, it started conflating VPNs, ZTNA and SDP as similar solutions while insinuating they share common security controls. That isn’t great messaging for a company that wants to replace VPNs with ZTNA because it thinks VPNs are the definition of insecure remote access. Also, is this with creativity maxed out? Where’s the poetry, the metaphor, the emotional resonance?!
An apple a day keeps the ghosts away
As I sigh with tentative relief, I’ll give full disclosure. I’m not above using Grammarly to catch minor errors that my eyes glaze over after working on the same article for hours. I actually enjoy Grammarly’s reassuring adage in the corner of my editing window: “To err is human. To edit is divine.” Big mood, as the kids say. But even Grammarly gets confused. Still, that doesn’t mean ChatGPT isn’t a good starting point to get something on the page.
PR clients are already asking about AI writing tools, so Engage PR will monitor the model’s progress and its potential to boost clients’ programs. The tool still depends on human intelligence to input a suitable prompt and fact-check the output. It pulls from what’s already online, so it may leave you with extensive editing if you’re a startup with highly specific messaging trying to reinvent the wheel. When it comes to technical or creative content, it seems that humans are outrunning the ghosts in the machine. Time will tell if the ghosts catch up. But Pac-Man and Ms. Pac-Man seem safe in the maze, for now.
.png)
.png)
